Case-Based Consumer Theory

The neo-classical theory of consumer behavior, while a powerful tool, suffers from some well-known flaws. Specifically, it assumes a highly, often unrealistically rational utility-maximizing consumer and sheds little light on the dynamic nature of consumption decisions. In this paper we make some preliminary steps towards an alternative theory of consumer choices, restricted to the case of repeated "small" decisions. We assume that the consumer is choosing among products, rather than bundles, and that (s)he is a "case-based decision maker." In particular, such a consumer is not necessarily "optimizing" and may be "satisficing' in the sense of March and Simon (1958). The aggregation of choices among products implicitly defines a choice of a "bundle." It turns out that if the "aspiration level" of the consumer is relatively low, (s)he tends to be satisficed and may choose a "corner" solution, which is not necessarily "utility Maximizing" in the classical sense. If, however, the aspiration level is relatively high, the consumer keeps switching among the products, and their relative frequencies converge to an interior point in the bundles space, as suggested by the classical theory under the assumption of convex preferences. Furthermore, in our model we find that the "utility" of a product is closely related to the (limit) relative frequency with which it is consumed. Thus this model offers a new definition of a product's "utility," as a cardinal measure of desirability. To study the effect of changes in market conditions, we propose to incorporate a product's price directly into its utility. According to this view, the consumer does not consider the "utility" of each product (or bundle) as separate from the budget constraint. Rather, the fact that a certain product is expensive is implicitly assumed to alter the experience of consuming it. It follows that the consumer's reaction to price changes is "immediate," and does not require to (implicitly) solve the new optimization problem. Moreover, such a consumer may well respond to price changes without necessarily maximizing his/her utility subject to the budget constraint. Next we study the relationship between the weight attached to the price in the (linear) evaluation of a product and the budget constraint. We show that, under certain reasonable conditions, the total expenditure is a decreasing function of this parameter and it follows that there exists a unique value for this parameter which balances the consumer's budget. Finally, we introduce the notion of the "potential" of the utility function, which is akin to the neo-classical utility function. We model substitution and complementarity effects, which are related to the cross derivatives of the potential.